Merging ArtNet universes in OLA

[discop...@me.com]

Hi there, I've just started using OLA, and I want to sound off a few plans I have.

First, my setup. I have a lighting rig that currently spans 1 DMX universe. I control this using MaxMSP on my Mac and a single DMX USB PRO (http://www.enttec.com/?main_menu=Products&pn=70304) and the dmxusbpro external, and all works fine.

I am upgrading my rig soon and it will span at least 3 DMX universes. I'm buying a DMX King eDMX4 to increase my total universes available. (http://dmxking.com/artnetsacn/edmx4-tx-rdm)

There is another lighting engineer I work with, with his own lighting fixtures, and his own set of control gear (including a Chamsys MagicQ PC Wing, which outputs ArtNet - http://www.chamsys.co.uk/pcwingcompact).

We would like to combine our lighting rigs into one rig, so that we can both control it from our respective control gear - sometimes simultaneously. We should be able to do this by merging each of our universes using the HTP rule. Merging multiple universes using dedicated DMX512 hardware could be costly and clunky, so I'm trying to find out if it's possible to merge ArtNet universes (which I'm sure it must be) and whether this can be done in OLA (not so sure).

So far I've installed OLA on my Mac and set up ArtNet universes on my WiFi interface - 4 in, and 4 out. I can send data through the out ports using the Max OlaOutput external.

Can I somehow get my lighting buddy to send his universes to my OLA in ports, and merge them with the universes I am generating, before sending them onwards to the out ports?

[Jason Kyle]

The eDMX4 will take care of the merging for you already. It will accept 2 incoming streams and merge into 1 DMX output with either HTP or LTP on a per port basis.

We’ve just added sACN support in the eDMX firmware too and you can actually merge an Art-Net and an sACN stream together if that became a scenario for you in the future.

Doesn’t MaxMSP also have Art-Net support already? I’m pretty sure there’s a Java external and someone also wrote a native C external. That would eliminate the USB DMX box and then all merging is handled by the eDMX4.

 

Best Regards,

 

Jason Kyle

DMXking.com / JPK Systems Limited

--
The Open Lighting Group: open-l...@googlegroups.com, #openlighting (irc.freenode.org)
To unsubscribe from this group, send email to open-lightin...@googlegroups.com
For more options, visit https://groups.google.com/groups/opt_out?hl=en

[discop...@me.com]

Hi Jason, thanks for your reply. I ordered your eDMX4 yesterday :)

That's great that the eDMX4 can do the merging for me.

However, I'm still interested to know how OLA would handle ArtNet merging internally.

One scenario would be using all 4 universes on the eDMX4 *and* the 1 universe on the DMX USB PRO, giving me a total of 5 DMX output ports. I would want all 5 universes to be merged before they left OLA, as the DMX USB PRO does not do any merging of its own.

[Jason Kyle]

Take a look here:

http://www.opendmx.net/index.php/OLA_Merging_Algorithms

[discop...@me.com]

I've seen that page, but it doesn't seem to explain what I need.

Simpy put, I want to use OLA to merge two Artnet input universes to one Artnet output universe.

Is this possible? If so, what would the OLA patching settings look like?

This is what I've done so far:

patrick$ ola_dev_info
Device 1: ArtNet [192.168.0.8]
  port 0, IN ArtNet Universe 0:0:1, priority 100, patched to universe 1, RDM supported
  port 1, IN ArtNet Universe 0:0:2, priority 100, patched to universe 2, RDM supported
  port 2, IN ArtNet Universe 0:0:3, priority 100, patched to universe 3, RDM supported
  port 3, IN ArtNet Universe 0:0:4, priority 100, patched to universe 4, RDM supported
  port 0, OUT ArtNet Universe 0:0:11, patched to universe 11, RDM supported
  port 1, OUT ArtNet Universe 0:0:12, patched to universe 12, RDM supported
  port 2, OUT ArtNet Universe 0:0:13, patched to universe 13, RDM supported
  port 3, OUT ArtNet Universe 0:0:14, patched to universe 14, RDM supported

I think what I want to do is take the input of universe 1 and 2, merge them, and send them out on universe 11. Am I on the right track?

[Peter Newman]

I think, and this is guesswork, I've not tried it, that you want to add multiple input ports to a universe. It might be worth experimenting with the USB DMX Pro and ArtNet as input to an OLA universe, as I've a vague feeling that ArtNet universes are tied to OLA universes, so the second ArtNet universe in an OLA universe would end up on the same universe ID. But that's all guesswork.

[discop...@me.com]

Thanks Peter.

Ok, here's an example of taking the dmxusbpro input and one artnet input, (presumably) merging them, and sending them out on the dummy device output:

(output truncated)

patrick$ ola_dev_info
Device 1: ArtNet [192.168.0.8]
  port 0, IN ArtNet Universe 0:0:1, priority 100, patched to universe 1, RDM supported

Device 2: Dummy Device
  port 0, OUT Dummy Port, patched to universe 1, RDM supported
Device 10: Enttec Usb Pro Device
  port 0, IN Serial #: xxxxxxxx, priority 100, patched to universe 1

But trying to add any more Artnet inputs or outputs on the same universe fails - both in the web interface, and using the ola_patch command. And that would make sense if Artnet and OLA universes are tied - which they do appear to be.

So it would seem that merging 2 Artnet inputs to 1 Artnet output is not possible within OLA alone. If I could route the data through another program then perhaps I could merge there instead.

[Simon Newton]

The artnet port will merge two sources for the same universe. It can
either be HTP or LTP mode (configureable with ola_artnet).

The output of the merge will be combined with any other sources to
produce the universe output which can then be send to the USB Pro.

Simon

[discop...@me.com]

Hi Simon, thanks for your reply.

How about sending onwards to an Artnet output? So, 2 Artnet universes in, merged to 1 Artnet universe out. Is this possible?

Patrick

[Simon Newton]

On Wed, Aug 21, 2013 at 7:27 AM, <discop...@me.com> wrote:
> Hi Simon, thanks for your reply.
>
> How about sending onwards to an Artnet output? So, 2 Artnet universes in,
> merged to 1 Artnet universe out. Is this possible?

No, right now ArtNet universes are tied to OLA universes (if you bind
an ArtNet port to universe 1, it automatically receives on ArtNet
universe 1). There was talk of de-coupling them but that hasn't
happened yet.

Simon

[discop...@me.com]

Thanks for clarifying that Simon. I'll be keeping an eye on developments.

Patrick

[Peter Newman]

In which case, more guesswork/pondering, what about if you route it out as E1.31/sACN, then back into OLA as that to then transmit as ArtNet, i.e. use another network protocol to work around the issue.

[Disco Patrick]

Interesting workaround Peter. Are you saying that sACN universes are not tied to OLA universes?

[Peter Newman]

It seems I'm wrong again, it looks like OSC is the only thing that would currently let you uncouple/link universes, everything else is tied to the OLA universe you choose. I'm not sure how efficient OSC would be though if wrapped around.

[Taine]

The idea of decoupling in very general terms seems quite powerful. You would then be able to build "buffers" from associated inputs. This step could handle various input priority schemes and possible an variant on input patch.

 

The next step would be to build outputs from associated "buffers". This step would allow for the priorities and patching on the output side.

 

So you could build a "buffer" from two E1.31 universes based on their priority. You could then combine this "buffer" with a "buffer" from a Artnet input using HTP... and send this out as Shownet netslots 1001 thru 1512(Remembering that Shownet doesn't really case about universe boundries).

[Andrew Frazer]

This "decoupling" is done in Madrix ( www.madrix.com ) in its patching. They have "madrix" Universes and " dmx universes ". It does confuse a lot of people, that they are not the same! However the feature is very helpful, because you get a lot of flexibility in how devices can be configured.

> The idea of decoupling in very general terms seems quite powerful. You would then be able to build "buffers" from associated inputs. This step could handle various input priority schemes and possible an variant on input patch.
>
> The next step would be to build outputs from associated "buffers". This step would allow for the priorities and patching on the output side.

Its an interesting concept.

If you did this, you could possibly, then have 'processor' loops on universes as well. Much like a "FX loop" on a audio mixer. You could send a universe to a virtual effects processor and do something to it, before sending it back out..


For example, if you wanted to hard limit all dmx values to a maximum value. ( say 200 ), you could pass it through a "limiter", or if you wanted to reorder the data, ( chinese cheap RGB pixels are anot always RGB, they sometimes are BGR, or GBR etc etc ) .. you could have a 're-order-er'..

if anyones familiar with JackD, the concepts not that dissimilar.

[Simon Newton]

I've been kicking around this idea for years now. See the notes in
TODO from 2005 :).

I'm mostly holding off until we have a better idea of what E1.33 is
going to look like since that will influence the design.


Simon

[Andrew Frazer]

would e1.33 really influence the design? Part of the magic of OLA is that it seems to be "above" any protocols.
I guess e1.33 might provide a "control" between the different "light-fx" boxes even within a single box.

[Simon Newton]

On Sat, Sep 7, 2013 at 4:55 PM, Andrew Frazer
<andrew...@stellascapes.com> wrote:
>
> would e1.33 really influence the design? Part of the magic of OLA is that it seems to be "above" any protocols.
> I guess e1.33 might provide a "control" between the different "light-fx" boxes even within a single box.

It does because E1.33 specifies PIDs for enumerating endpoints (ports)
and patching. The internal model needs to be compatible with that.


Simon

[Andrew Frazer]

Pity that the e1.33 standard is becoming a political football then.